Talk to humans, do not start with a survey

Talk to humans, do not start with a survey

The instinct when founders decide to "do customer research" is to build a survey, blast it to a list, and wait for a spreadsheet to tell them what to do. It almost never works, and it teaches you the wrong lesson about what research is. Surveys are good at confirming what you already suspect at scale. They are terrible at telling you what you do not yet know to ask, which is the entire reason you are researching in the first place.

Real customer research starts with conversations. You sit with one buyer, you ask open questions, and you shut up. The value is in the words they choose, the order they tell their story in, the moment their voice changes when they hit the thing that actually hurt. A survey flattens all of that into a multiple-choice average. A conversation hands you the raw material your marketing, your product and your sales calls are built from.

The reason this matters so much in B2B is that your market is small and your buyers are specific. You are not optimising a checkout funnel for a million strangers. You are trying to understand a few hundred companies and the handful of people inside each who decide. At that scale, ten good conversations beat a thousand survey responses, because depth is what you are short of, not volume.

So the right order is conversations first, surveys later. Interview until you keep hearing the same handful of stories, then, if you need to size or rank what you heard, write a survey to quantify it. The survey validates the pattern. The interview finds it. Reverse that order and you confirm your own assumptions with great statistical confidence.

The trap to name out loud is that conversations feel slow and surveys feel efficient. That efficiency is an illusion when the output is wrong. Five hours of interviews that change your positioning are worth more than five hundred survey responses that decorate a slide.

INTERVIEW EWOUD: Tell the story of a time a customer conversation gave you something a survey never would have. What was the line or moment that shifted how you sold the product?